INDEX

Explanations

references to experimentation and innovation

oai_token-act-pair · gpt-4o-mini Triggered by @bot

experiment with

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 betale

-0.52

 icônes

-0.50

 Coat

-0.49

новниш

-0.48

følgelig

-0.48

AddTagHelper

-0.47

 Floors

-0.47

notify

-0.47

DataBinding

-0.46

requirement

-0.46

POSITIVE LOGITS

 experimentation

0.98

 experimenting

0.98

 experiments

0.90

 Experiment

0.82

 experiment

0.81

experiment

0.81

 EXPERIMENTS

0.79

Experiment

0.78

 experimented

0.78

experiments

0.77

Activations Density 0.023%

references to experimentation and innovation

experiment with

No Comments

No Known Activations